MyD88 signalling in B cells and antibody responses during Oropouche virus-induced neurological disease in mice

Abstract

Background: Oropouche virus (OROV) is a neglected insect-borne orthobunyavirus that causes a febrile illness, neurological disease, and pregnancy complications in humans across an endemic area spanning South and Central America. The host factors associated with disease pathogenesis have nonetheless remained obscure, and little is known about the immune determinants of protection against OROV.

Methods: We tracked morbidity, mortality, viral loads, and serum neutralisation in wild-type (WT), Rag1^-/-, CD19-Cre⁺Ifnar^fl, and CD19-Cre⁺MyD88^fl mice and performed immunophenotyping experiments, passive serum transfers, and adoptive cell transfers to determine how early antibody responses and B cell subsets control viral replication and dissemination to the central nervous system after infection with OROV.

Findings: In line with a protective role for B cells, WT mice efficiently produced OROV-specific antibodies within 6 days of infection. Serum transfer containing neutralising IgM from WT to Rag1^-/- mice prevented neurological disease in OROV-challenged mice. CD19-Cre⁺MyD88^fl mice but not CD19-Cre⁺Ifnar^fl mice were vulnerable to neurological disease and produced lower titres of OROV-specific antibodies that exhibited suboptimal neutralisation and potency compared with MyD88-sufficient mice. CD19-Cre⁺MyD88^fl mice also presented with reduced numbers of marginal zone B (MZB) cells and plasmablasts after infection, which were associated with high viral burdens and lethality. Adoptive transfer of MZB cells from WT mice protected CD19-Cre⁺MyD88^fl mice and partially protected Rag1^-/- mice from lethal infection with OROV.

Interpretation: Early MyD88 signalling in B cells is required for optimal antibody responses that limit viral replication and neurological disease in mice infected with OROV.

Funding: São Paulo Research Foundation (FAPESP), National Council for Scientific and Technological Development (CNPq), Coordination for the Improvement of Higher Education Personnel (CAPES), Unicamp Research Affairs Office, PIPAE University of São Paulo, Wellcome Trust, and National Institute of Science and Technology on Photonics Applied to Cell Biology (INFABIC, Unicamp).

Keywords: Emerging viruses; IgM; Innate immune response; Marginal zone B cells; Oropouche virus; Vector-borne diseases.

Fig. 1

Mortality, weight loss, morbidity, and viral loads for Rag1^−/− mice after OROV infection. (a) Survival and (b) weight of C57BL/6 WT mice aged 4–6 weeks (grey, n = 14), Rag1^−/− mice aged 4–6 weeks (red, n = 14), and Rag1^−/− mice aged 9–12 weeks (orange, n = 14) over 21 days post-infection (dpi) with 10⁵ FFU of OROV. Statistical differences were assessed using a log-rank test with Bonferroni correction (a). ∗∗∗∗p < 0.0001. (c) Clinical scores for Rag1^−/− mice aged 4–6 weeks (n = 14) after OROV infection. Bars indicate the percent occurrence for each clinical score among all mice. Clinical scores were defined as follows: (1) active animal, no disease; (2) slow walking, slightly closed eyes, or slight impairment of balance; (3) apathy, ataxia, closed eyes, severe impairment of balance, circular walking, lethargy, or paralysis; or (4) moribund or deceased animal. (d–j) Viral loads in serum and tissues from infected Rag1^−/− mice (red, n = 8 at each time point) and WT mice (grey, n = 8 at each time point) measured via focus-forming assay (circles correspond to left y-axis) or RT-qPCR (squares correspond to right y-axis). Statistical differences were assessed using the Kruskal–Wallis test with Dunn’s post-hoc test for each group versus day 1 post-infection. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001, ns = not significant. Symbols represent death events (a), mean ± SEM (b), or individual viral loads (d–j), and lines represent mean values (d–j). Data were pooled from at least three independent experiments. FFU Eq = focus-forming unit equivalents; L.O.D. = limit of detection.

Fig. 2

Mortality, weight loss, morbidity, and viral loads for μMT mice after OROV infection. (a) Survival and (b) weight of C57BL/6 WT mice (grey, n = 14), μMT mice (orange, n = 18), and TCRβδ mice (green, n = 21) over 21 days post-infection (dpi) with 10⁵ FFU of OROV. Statistical differences were assessed using a log-rank test with Bonferroni correction (a). ∗∗∗∗p < 0.0001. (c) Clinical scores for μMT mice (n = 18) after OROV infection. Bars indicate the percent occurrence for each clinical score among all mice. Clinical scores were defined as in Fig. 1. (d–j) Viral loads in serum and tissues from infected μMT mice (n = 10 at each time point) measured via focus-forming assay. Statistical differences were assessed using the Kruskal–Wallis test with Dunn’s post-hoc test across all days post-infection. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001, ns = not significant. Symbols represent death events (a), mean ± SEM (b), or individual viral loads (d–j), and lines represent mean values (d–j). Data were pooled from at least three independent experiments. L.O.D. = limit of detection.

Fig. 3

Mortality, weight loss, morbidity, and viral loads for CD19-Cre⁺MyD88^fl mice after OROV infection. (a) Survival and (b) weight of CD19-Cre⁺MyD88^fl/fl mice (blue circles, n = 19), CD19-Cre⁺MyD88^fl/wt mice (blue triangles, n = 11), and Cre^–MyD88^fl/fl or CD19-Cre⁺MyD88^wt/wt mice (grey circles, n = 17) over 21 days post-infection (dpi) with 10⁵ FFU of OROV. Statistical differences were assessed using a log-rank test with Bonferroni correction (a). ∗p < 0.05. (c, d) Clinical scores for WT mice (n = 19) and CD19-Cre⁺MyD88^fl mice (n = 19) after OROV infection. Bars indicate the percent occurrence for each clinical score among all mice. Clinical scores were defined as in Fig. 1. (e–k) Viral loads in serum and tissues from infected CD19-Cre⁺MyD88^fl mice (blue, n = 12 at each time point) and Cre^–MyD88^fl/fl or CD19-Cre⁺MyD88^wt/wt mice (grey, n = 12 at each time point) measured via focus-forming assay. Statistical differences were assessed using the Kruskal–Wallis test with Dunn’s post-hoc test for each group versus day 1 post-infection. ns = not significant. Symbols represent death events (a), mean ± SEM (b), or individual viral loads (e–k), and lines represent mean values (e–k). Data were pooled from at least three independent experiments. L.O.D. = limit of detection.

Fig. 4

Mortality, weight loss, and viral loads for Rag1^−/− mice and IgM responses after OROV infection. (a) Survival and (b) weight of Rag1^−/− mice inoculated on the day of OROV challenge with sera harvested from WT mice 3–6 days post-infection (dpi) (coloured numbers, n = 6 per group). (c) Viral loads in serum and tissues from mice in (a, b) measured via focus-forming assay (coloured numbers, n = 6 per group). (d) Survival and (e) weight of Rag1^−/− mice inoculated on the day of mock manipulation (orange, n = 5) or OROV challenge (red, n = 7) with β-ME-treated sera harvested from WT mice on day 6 post-infection. (f) Viral loads in serum and tissues from mice in (d, e) measured via focus-forming assay (n = 5 for mock manipulation, n = 7 for OROV challenge). Statistical differences were assessed using a log-rank test with Bonferroni correction (a, d) or the Kruskal–Wallis test with Dunn’s post-hoc test across all tissue samples (c, f). ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001, ns = not significant. Symbols represent death events (a, d), mean ± SEM (b, e), or individual viral loads (c, f), and lines represent mean values (c, f). Data were pooled from at least three independent experiments. (g) Anti-OROV ELISA and (h) FRNT analyses of sera harvested from WT mice on days 6 or 14 post-infection before (grey, n = 10) and after treatment with β-ME (red, n = 5). (i) FRNT₅₀ values for individual mice in (g, h) (n = 10 per group). Statistical differences were assessed using the Mann–Whitney test with Bonferroni correction (g, i). ∗p < 0.05, ∗∗∗∗p < 0.0001, ns = not significant. Symbols represent individual mice (g, i) or mean ± SEM (h), bars represent mean ± SEM (g, i), and lines represent three-parameter log-logistic regression curves for relative infection per dilution (h). Data were pooled from at least three independent experiments. L.O.D. = limit of detection.

Fig. 5

Antibody responses in mice with MyD88-deficient B cells during OROV infection. (a, b) Quantification and (c, d) avidity of OROV-specific IgM (a, c) and IgG isotypes (b, d) in serum samples (10-fold dilution) from WT mice (grey, n = 8 per condition/time point) and CD19-Cre⁺MyD88^fl mice (blue, n = 8 per condition/time point) after mock manipulation (NI = noninfected) or at the indicated days post-infection (dpi) with 10⁵ FFU of OROV. (e) Neutralisation potency index (log FRNT₅₀/Ig) of serum samples from WT mice and CD19-Cre⁺MyD88^fl mice (n = 8 for 6 dpi and n = 10 for 14 dpi). Statistical differences were assessed using the Mann–Whitney test with Bonferroni correction (a–e). ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001, ns = not significant. Symbols represent individual mice, and bars represent mean ± SEM (a–e). (f–h) FRNT analysis of sera harvested from (f) WT mice (grey, n = 6 for NI, n = 6 for 6 dpi, and n = 16 for 14 dpi) and (g) CD19-Cre⁺MyD88^fl mice (blue, n = 3 for NI, n = 8 for 6 dpi, and n = 8 for 14 dpi) after mock manipulation or at the indicated days post-infection, with (h) FRNT₅₀ values for individual WT mice (grey, n = 5 for NI, n = 6 for 6 dpi, and n = 16 for 14 dpi) and CD19-Cre⁺MyD88^fl mice (blue, n = 3 for NI, n = 7 for 6 dpi, and n = 7 for 14 dpi). Ascitic fluid (anti-OROV) was used as a positive control (purple). Statistical differences were assessed using the Mann–Whitney test with Bonferroni correction (h–i). ∗∗p < 0.01, ns = not significant. Symbols represent mean ± SEM (f, g) or individual mice (h), lines represent three-parameter log-logistic regression curves for relative infection per dilution (f, g), and bars represent mean ± SEM (h). (i) FRNT₅₀ values for individual CD19-Cre⁺MyD88^fl mice that survived (blue, n = 3) or succumbed to the infection (red, n = 2). Statistical differences were assessed using the Mann–Whitney test with Bonferroni correction (i). ns = not significant. Symbols represent individual mice, and lines represent mean values (i). nc = noncollected sample. (j) Survival and (k) weight of Rag1^−/− mice inoculated on the day of OROV challenge with sera harvested from CD19-Cre⁺MyD88^fl mice on day 6 (circles, n = 9) or day 14 post-infection (squares, n = 7). (l) Viral loads in serum and tissues from mice in (j, k) measured via focus-forming assay (n = 5 per group). Statistical differences were assessed using a log-rank test with Bonferroni correction (j) or the Kruskal–Wallis test with Dunn’s post-hoc test across all tissue samples (l). ns = not significant. Symbols represent death events (j), mean ± SEM (k), or individual viral loads (l), and lines represent mean values (l). L.O.D. = limit of detection.

Fig. 6

B cell subset development and maintenance in MyD88-deficient mice during OROV infection. (a) Uniform Manifold Approximation and Projection (UMAP) representation of splenic CD3^– Gr1^– F4/80^– lymphocytes from WT mice (grey, n = 8 per condition/time point) and CD19-Cre⁺MyD88^fl mice (blue, n = 3 for noninfected [NI] and n = 8 at each time point) after mock manipulation or at days 6 or 14 post-infection (dpi) with 10⁵ FFU of OROV. (b) B cell subset frequencies among splenic CD3^– Gr1^– F4/80^– lymphocytes from (a). (c–m) B cell numbers per spleen for the indicated subsets from WT mice (grey, n = 8 per condition/time point) and CD19-Cre⁺MyD88^fl mice (blue, n = 3 for NI and n = 8 at each time point) after mock manipulation or at days 6 or 14 post-infection with 10⁵ FFU of OROV. B cell subsets were evaluated via flow cytometry as (c) B220^–, (d) B220^int, (e) B220^hi, (f) B220^– CD138⁺ (plasma cells), (g) B220^hi CD23^– (marginal zone B cells), (h) B220^hi CD23⁺ (follicular B cells), (i) B220^hi CD138^– GL7⁺ CD38^– (germinal centre B cells), (j) B220^hi CD138^– GL7^– CD38⁺ (memory B cells), (k) B220^hi CD138^int CD38⁺ (plasmablasts 1), and (l) B220^int CD138⁺ (plasmablasts 2), alongside total cellularity (m). Statistical differences were assessed using the Mann–Whitney test with Bonferroni correction. ∗p < 0.05, ∗∗p < 0.01, ns = not significant. Symbols represent individual mice, and bars represent mean ± SEM. Red symbols indicate samples harvested from mice with clinical scores of 2 (at 6 dpi) or 3 (at 14 dpi).

Fig. 7

The role of marginal zone B cells during OROV infection. Factor analysis of mixed data (FAMD) using mouse genotypes, IgM and IgG quantity, avidity, and neutralisation potency (indices and FRNT₅₀ values), viral loads in brains, death events, and the cellularity of B cell subsets. (a) FAMD factor map showing the categorical variable of genotype for WT mice (n = 8 for noninfected [NI], n = 8 for 6 dpi, and n = 10 for 14 dpi) and CD19-Cre⁺MyD88^fl mice (n = 7 for NI, n = 11 for 6 dpi, and n = 10 for 14 dpi). Colour-coded circles represent individual mice, white circles represent group centroids, and crossed circles represent the categorical variables of live (middle) and death events (top). (b) Quantitative variables are represented by arrows coloured to indicate relative contribution in each case (key) for WT mice (n = 8 for NI, n = 8 for 6 dpi, and n = 10 for 14 dpi) and CD19-Cre⁺MyD88^fl mice (n = 7 for NI, n = 11 for 6 dpi, and n = 10 for 14 dpi). (c, d) The percent contribution of each variable to (c) Dim-1 and (d) Dim-2 for WT mice (n = 8 for NI, n = 8 for 6 dpi, and n = 10 for 14 dpi) and CD19-Cre⁺MyD88^fl mice (n = 7 for NI, n = 11 for 6 dpi, and n = 10 for 14 dpi). The red line indicates the expected average value for a uniform contribution. (e) Correlation matrix colour-coded for percent contribution and variables for each dimension for WT mice (n = 8 for NI, n = 8 for 6 dpi, and n = 10 for 14 dpi) and CD19-Cre⁺MyD88^fl mice (n = 7 for NI, n = 11 for 6 dpi, and n = 10 for 14 dpi). Data were pooled from at least three independent experiments. Plasmablasts 1 = B220^hi CD138^int CD38⁺ cells, plasmablasts 2 = B220^int CD138⁺ cells. (f) Survival of Rag1^−/− mice (red circles, n = 6) and CD19-Cre⁺MyD88^fl mice (light grey squares, n = 6) after adoptive transfer of WT MZB cells or Rag1^−/− mice (blue circles, n = 5) and CD19-Cre⁺MyD88^fl mice (dark grey circles, n = 5) after adoptive transfer of CD19-Cre⁺MyD88^fl MZB cells over 21 days post-infection (dpi) with 10⁵ FFU of OROV. Statistical differences were assessed using a log-rank test with Bonferroni correction (f). ∗∗p < 0.01, ns = not significant (f). Symbols represent death events (f). (g) FRNT analysis of sera from CD19-Cre⁺MyD88^fl mice at terminal harvest after adoptive transfer of WT MZB cells (light grey, n = 6) or CD19-Cre⁺MyD88^fl MZB cells (dark grey, n = 5) during OROV infection as in (f). Symbols represent mean ± SEM, and lines represent three-parameter log-logistic regression curves for relative infection per dilution (g). (h) FRNT₅₀ values for individual mice in (g) after adoptive transfer of WT MZB cells (light grey, n = 6) or CD19-Cre⁺MyD88^fl MZB cells (dark grey, n = 5). Statistical differences were assessed using the Mann–Whitney test (h). ∗∗p < 0.01. Symbols represent individual mice, and bars represent mean ± SEM (h). MZB = marginal zone B.

Fig. 8

Interpretation of the study. Early antibody responses produced mainly by MZB cells in a MyD88-dependent manner restrict viral replication and spread to the central nervous system, protecting mice from encephalitis and subsequent death after infection with OROV.

Supplemental Fig. S1